Fan system monitoring via cloud service
Kimmo Huoman · LUTPub (LUT University) · 2016
The majority of electricity consumption in industrial and service sectors in the European Union is caused by electric motors, largely by fluid handling systems such as fans, pumps, and compressors. By utilising energy efficient motors and variable speed drives, the energy savings potential in these sectors is estimated to be almost 90 TWh per year, even with economically viable investments. In relation, this is more than the electricity consumption of Finland. Most of the potential savings are achieved by replacing valves, dampers, and other flow restricting devices in fluid handling systems with a rotational speed control using a variable speed drive. The variable speed drive can serve as more than just a method of adjusting the rotational speed. It can provide an excellent, centralised source of information. Recent studies show that the variable speed drive can provide values for estimating multiple features of fluid handling systems, features currently being monitored by dedicated sensors. By reducing the amount of sensors and moving on to sensorless estimation methods, the system complexity can be reduced. Currently variable speed drive monitoring is typically implemented only to applications where the uninterrupted operation is crucial either for safety or minimising production losses. However because of effects caused by the Internet of Things, the cost and complexity of connecting hardware devices to the Internet is constantly decreasing. By connecting variable speed drives to a cloud service, both the available storage space and processing power are practically unlimited. In this master’s thesis, a cloud service capable of retrieving data from variable speed drives is developed. A fan monitoring application is built on top of the stored data, capable of estimating features such as the produced airflow. This is achieved by reviewing the current methods for operating point estimation based on mathematical model and the manual work required to obtain the equations used in the models. As an end result, a prototype cloud service is deployed with fan system monitoring features.